向家乡投资：信息、熟悉还是代理问题？

Invest to the Hometown: Information, Familiarity or Agency Problem?

本文探讨公司投资中是否存在CEO的"家乡偏好",并且在信息、熟悉假说之外,基于身份认同经济学进一步提出代理问题假说,以检验中国情景下公司投资的家乡偏好动机。以手工整理的民营上市公司CEO家乡以及母、子公司地理数据,本文通过民营上市公司—地区—年度样本以及仅包含了异地任职CEO公司的子样本实证均发现:公司会更多在CEO的家乡建立异地子公司,该结论在经过多种稳健性测试之后仍然成立;并且CEO家乡的子公司相比其他异地子公司绩效更差。这说明公司投资中确实存在CEO的"家乡偏好",并且其更适合以代理问题假说来解释,而非信息和熟悉假说。进一步研究发现,CEO自身代理问题越轻(CEO持股、年轻CEO)和外部治理越强(机构投资者持股高、分析师跟踪多)均会抑制公司向CEO家乡的投资;代理问题假说解释的"家乡偏好"投资并不存在于董事长个体中,我们虽然也发现公司会更多向董事长家乡投资,但这种投资的绩效并无显著差异,相对更适合以熟悉假说解释。本文从家乡偏好视角将身份认同引入经理人代理行为,不仅在公司投资层面丰富了家乡偏好的经济后果,也以经理人个人偏好拓展和丰富了经理人代理问题的相关研究,提供了中国情景下经理人代理问题的新证据。

This study investigates whether CEOs exhibit home bias in corporate investment. Based on identity economics, we put forward agency conflicts hypothesis beyond the information and familiarity hypothesis to test the motivation of home bias in corporate investment under the background of China. We manually search for hometown of CEOs in private listed companies and geographic data of the parent company and its subsidiaries. Our sample includes both state-year sample of private listed companies and subsample of companies which CEOs work non-locally. We find that firms are likely to establish more subsidiaries in CEOs' hometown and this conclusion is valid. Besides, subsidiaries in CEOs' hometown underperform other nonlocal subsidiaries, suggesting that home bias does exist in corporate investment and it is largely driven by agency issues instead of the information and familiarity. Furthermore, we show that home biased investments will be suppressed in the companies with better external governance(higher institutional investors holdings,more analysts tracking) and weaker agency problems on CEOs(CEO holdings, young CEO), which also indirectly support agency conflicts hypothesis. This paper introduces identity in the managers' agent behavior from the perspective of home bias, thus not only enriching the economic consequences of home bias in the aspect of corporate investment, but also expanding the research on agency problem with the managers' individual preference. Its findings help to provide new evidence for agency problem in China.